https://github.com/willguimont/onnxstream-docker
Run OnnxStream in Docker
https://github.com/willguimont/onnxstream-docker
Last synced: 10 months ago
JSON representation
Run OnnxStream in Docker
- Host: GitHub
- URL: https://github.com/willguimont/onnxstream-docker
- Owner: willGuimont
- License: mit
- Created: 2023-09-11T16:37:12.000Z (almost 3 years ago)
- Default Branch: master
- Last Pushed: 2024-04-28T01:23:44.000Z (about 2 years ago)
- Last Synced: 2025-04-05T03:02:44.436Z (over 1 year ago)
- Language: Dockerfile
- Size: 2.93 KB
- Stars: 5
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Dockerfile for OnnxStream
This is a Dockerfile to run [OnnxStream](https://github.com/vitoplantamura/OnnxStream) to run Stable Diffusion on less powerful computers.
## Build
```bash
docker build -t onnxstream .
```
## Run
```bash
mkdir output
docker run --rm -it \
--mount type=bind,source=$(pwd)/output,target=/app/out \
onnxstream:latest sd --models-path /weights/SD --output /app/out/out.png --prompt "a penguin standing in a forest" --neg-prompt "ugly, blurry"
# Parameters
# --models-path Sets the folder containing the Stable Diffusion models. Weights are extracted to /weights/SD in the Docker container
# --ops-printf During inference, writes the current operation to stdout.
# --output Sets the output PNG file. This should be inside of a volume to make the file accessible to the host computer.
# --decode-latents Skips the diffusion, and decodes the specified latents file.
# --prompt Sets the positive prompt.
# --neg-prompt Sets the negative prompt.
# --steps Sets the number of diffusion steps.
# --save-latents After the diffusion, saves the latents in the specified file.
# --decoder-calibrate Calibrates the quantized version of the VAE decoder.
# --decoder-fp16 During inference, uses the FP16 version of the VAE decoder.
# --rpi Configures the models to run on a Raspberry Pi Zero 2.
```